Page Two articles are from our reader community. Sign up for an account.


There are many ways to improve service and provide the best bang for our transit dollars. You can improve or even remove low performing routes – that was essentially the focus on Metro’s September 2014 reductions. But, you can also tweak high performing routes and get either deliver increased service at the same cost or reduce service without reducing frequency. I’d like to take a stab at improving one of the most popular routes: the SoundTransit 550.

The 550 is responsible for 16% of SoundTransit’s boardings, with close to 10,000 rides a day. In rush hour, it runs as frequently as 11 times an hour. It is so popular, it is essentially being replaced with a light rail line in 2023. That’s still 8 years away, and there’s room to improve this busy route right now.

I propose replacing some of 550’s rush hour runs with a new 551 – a super express to Seattle. The 550 begins at the Bellevue Transit Center, makes a few downtown Bellevue stops and then goes down Bellevue Way to serve the S. Bellevue Park and onto Mercer Island, Rainier Freeway and Seattle. The great majority of people on this route are just going between Bellevue and Seattle and the intermediate stops just waste those people’s time and our transit dollars.

The 551 would start at Bellevue Transit Center, stop at one or two (though likely different than the 550) downtown stops and proceed directly to Seattle. That would allow the bus to get on the freeway at SE 8th St, using a queue jump and dedicated HOV exit lane to go directly from I405 to I90. Without traffic and stops this saves 4 minutes according to Google Maps), I suspect the actual savings are much larger. Skip the Mercer Island stop and the Ranier Freeway station and head directly to the tunnel and make the regular tunnel stops. Do the same on the return.

I’m not sure how many of the 550 runs could be converted to 551. You wouldn’t want to convert too many, as there are enough users of the South Bellevue Park & Ride and the western Bellevue stops near Bellevue Way who would miss out. But even if you convert just 1 trip per hour during rush hour, you improve the rides of hundreds of people and you might be able to squeeze out an extra run you couldn’t do otherwise.

Downsides? There are some. First, the confusion at least at first — it would be annoying if you got on a 551 bus looking for South Bellevue Park and ride and you end up in Seattle instead — though good outreach, driver announcements and signage could help. Also, you would get people used to the super express with no stops when the matching light rail service that opens in 8 years can’t match it. But light rail can also bypass the traffic the 550 gets stuck behind in Bellevue, so the effect might be a wash. And the traffic is just heating up — there a number of new towers going up in Downtown and the light rail construction will yield some serious traffic issues.

SoundTransit (and STB commenters), what do you say?

33 Replies to “Service Improvement Idea: The Super Express Bellevue to Seattle”

  1. I’m not sure how useful stops in downtown Bellevue that aren’t shared with the 550 would be for a service that runs just a handful of times per day on an irregular schedule. It’s not totally analogous, but… you might think a peak express route between Ballard and UW, bypassing Wallingford, would be a good idea, but hardly anyone used the old 46 trips that did this — I think largely because they didn’t stop in the same places as the 44.

    I don’t have numbers on this, but I have to believe South Bellevue P&R, which isn’t walkable to much and isn’t a major transfer point, sees very little counter-peak 550 usage (since downtown Bellevue’s parking situation isn’t difficult or expensive enough for people to drive as close as South Bellevue then park and take the bus, like people do along Lake Union in Seattle). A lot of express patterns are forward-peak only, but the 550 runs enough counter-peak trips that it could skip South Bellevue sometimes in the counter-peak without abandoning it, saving real passengers real time and delaying almost nobody. Or if ST at some point considers converting some deadheads into extra counter-peak service, a super-express would be less expensive than the whole 550 (though far from free).

    Of course, with South Bellevue P&R closing for Link construction fairly soon, this is a good time to consider whether the 550 should bypass that area when that happens.

    1. Serving South Bellevue has hardly any time penalty; the outbound stop is a pull-off from Bellevue Way, and the inbound stop requires a short jog into the bus bay (no loop-de-parking lot). Skipping it would save maybe 30 seconds outbound and a minute or two inbound. Adding a bus-only light for the left into the bay inbound would shave this down even more.

  2. I always thought the 550 should have gone on 405. However, when ST implemented the route it asked Bellevue which it preferred, and the majority of people said to keep some Bellevue Way stops. Of course, that was twenty years ago, and what people wanted then isn’t necessarily what they want now. But on the other hand Bellevue Way has densified greatly since then, and is in the midst of even more. So if you want regional transit next to blocks of multifamily housing, you’ve got it.

    Another downside is traffic on 405, which may be slower than Bellevue Way several times a day.

    People do use the South Bellevue P&R on weekends. On Sunday afternoons it’s sometimes standing room only westbound, although maybe only 1/4 or 1/3 of that is people from that stop. I can’t tell whether they parked/transferred/biked, but I assume 98% of them parked. It’s not a pleasant place to transfer to the 240, which is hourly on Sundays so not used much.

    1. 405 between I-90 and downtown Bellevue can be snot slow during peak. How well do the HOV lanes perform? Are they always faster than the current 550 route between I-90 and BTC?

      1. You’ll note I suggested getting on at SE8th instead of 6th. 6th puts you in the left hand HOV lanes which are very slow in rush hour. SE8th puts you in the I-90 Turn HOV Right Hand Lane which is a straight shot even during rush hour.

    2. The 240 doesn’t even go to South Bellevue P&R anymore – it goes to Eastgate instead. Transfer opportunities that still exist at South Bellevue P&R include the 560 (to a few office buildings along 112th or south to Newport Hills/Kennydale) and the 241 (to Factoria). Not a lot of people making those transfers.

      The current plan for the South Bellevue station construction is for the parking to go away for several years, but for buses to continue to stop there for transfer purposes and local neighborhood access. You could probably make at least half the peak-hour buses, if not more, skip that stop and no one would know the difference.

  3. I’d suggest that your additional stop in Bellevue should be at Wilburton P&R on the way to I-405. I’d also argue that the cost in travel time to serve MI P&R and Rainier Freeway Station are minimal and should be included on the route.

    1. Wilburton P&R is out of the way for the current 550 routing and why would you get on 405 at Wilburton when there are nice HOV ramps right at Bellevue a Transit Center?

      1. Remember that the exit to I-90 peels off the right, while the BTC HOV entrance goes to the left. There’re two HOV lanes between SE 8th and I-90, and the plan is for the 551 to use the right-side HOV lane which does exit onto 90. That does make it go right by Wilberton P&R.

      2. It can use the outer HOV lane which starts from SE 8th St on-ramp and connects to I-90 HOV Lane while going to Seattle. Using the NE 6th St ramps will require the bus to weave through 4 lanes of traffic to get to the exit to I-90.

        While coming from Seattle, there is a exit only lane at SE 8th St which generally moves a lot faster than the general lanes.

      3. My routing from BTC would be:

        NE 6th EB to 112th NE
        112th NE SB to Main St.
        112th SE SB to SE 6th
        SE 6th EB to 114th SE
        114th SE SB stop on-street at Wilburton P&R
        SB to I-405 SB

        As has been noted, there is an HOV lane on the right to get to I-90.

      4. Let me revise that. To use the same layover and BTC bay as the 550, it should go SB on 108th NE to Main St, then EB on Main to 112th SE.

  4. The South Bellevue P&R is going to close for five years while Link is built. All of that parking is likely to be dispersed across several smaller lots around Bellevue.

    This might be exactly the way to address that. Have the 551 get on the 405 as you suggest, and have the regular 550 run local around the replacement parking.

  5. I’ve often thought about this — I think it’s a very good idea. In fact I would make it even more agressive than you’ve suggested.

    Starting where the 550 currently does — leave the transit center, take a left of 108th, then a left on 4th St. Use the onramp to 405 S direct from 4th St (stay to the right), then head towards Seattle.

    On the return, when you get on 405 N, merge over the left of the highway and take the NE 6th St HOV exit. This returns you to the 550’s current bay where you do it all over again.

    A very simple service pattern. If the 550 already has 11 trips / hour at peak, make ~50% of them 551’s. The goal should be to always have 10 minute headways on the 550. Any routes above this then convert to 551’s.

  6. I don’t think this is a very good idea for multiple reasons.
    1) The 550 generally approximates the future East Link alignment from Seattle to Bellevue. You want to maintain a service pattern today that closely matches this. The 551 would break this up.
    2a) Mercer Island and South Bellevue together account for just as much weekday ridership as Bellevue Transit Center (seriously; go look at the 2014 SIP, which has stop level boarding information). Skipping these two places (along with Rainier) cuts out almost half of the 550’s ridership.
    2b) You wouldn’t actually save very many platform hours. A good chunk of the 550’s peak service is via single-trip runs, so these have significant deadheading. Saving 10 minutes of travel time while in-service is a small fraction over the total platform hours each peak trip requires.
    2c) Because of the above, any minimal savings in platform hours would be lost through decreased ridership, making the route much less cost effective.
    3) The slowest/most unreliable portion of the 550 is generally in downtown Bellevue, trying to cross NE 8th St to/from the terminal at the library and making the jog from Bellevue Way to the transit center; the turns onto/off of NE 4th St are horrible. In the PM peak, Bellevue Way can be an complete cluster, especially southbound, but 405 (and all the approaches to the ramps) is also horrible so not a great alternative. The 551 idea doesn’t really address the biggest source of delay and unreliability, instead saving a couple minutes by skipping stops in the (generally fast) I-90 corridor.
    4) The 550 has been built up as one of the major trunk routes. If you’re going to somewhere along the line, you get on. If you split off a 551, you’ve increased the complexity of the network.
    5) Creating a 551 relies on endpoint fallacy, focusing on the Bellevue and Seattle ends of the route at the expense of destinations along the way. While it is true that at peak the bus is largely serving commuters going all the way to one or the other, there are a not-insignificant number of on-offs at midpoints on the route.

    The last thing the 550 needs is a peak express overlay. Rather, it needs immediate BRT-lite infrastructure like off-board payment, TSP, queue jumps, etc., to address the areas causing the most problems and bridge the gap until East Link opens.

    1. it needs immediate BRT-lite infrastructure like off-board payment, TSP, queue jumps, etc., to address the areas causing the most problems and bridge the gap until East Link opens.

      I doubt there’s much appetite for putting a lot of capital into a service that’s going away in 2023. Even if it is the 550.

      1. That’s why I referred to BRT-lite features that generally won’t cost armloads of money. The 550 will never get dedicated lanes or any more grade separation than it already has, but things like TSP, queue jumps crossing NE 8th St, and a signal to enter South Bellevue P&R southbound could be implemented relatively inexpensively and would have a positive impact on travel time and reliability. Off-board payment is a different matter entirely, of course, since that is wrapped up in DSTT operations.

      2. If the 550 ends in 2023, what is going to replace it’s route between dt. Bellevue and South Bellevue P and Ride? I know there’s a KC route that basically mirrors that section of 550, but it is infrequent and I believe has very limited hours.

        I’m guessing they will instead just truncate 550 so it runs only from dt. Bellevue to South Bellevue. It would make no sense to get rid of it entirely. Stops along that route would go from having very good service frequency to having terrible service frequency. Hell, there is no other transit route even running along Main Street in Old Bellevue anymore.

        Even running 550 just from dt. Bellevue to South Bellevue would mean about half the buses currently used on 550 could be put towards more service on other routes. But eliminating frequent service along Bellevue Way in SE Bellevue would make no sense.

      3. I’d anticipate that’s the time to have a far-reaching realignment of Bellevue bus services. There would be enough available service hours to build out a lot of what Bellevue wants to see in terms of connections to Link, including along Bellevue Way.

      4. The 550 goes away when East Link opens.

        Based on the current network, there are several choices to serve Bellevue Way:
        1) Reroute the 249 so that it continues on Bellevue Way until SE 16th, then jogs over to 104th at that point. This removes a little bit of coverage on 104th between SE 8th and SE 16th in favor of maintaining coverage on Bellevue Way (the 550 has no stops on Bellevue Way south of SE 16th)
        2) Reroute the 241 from 108th to Bellevue Way. This deletes all coverage in Surrey Downs.
        3) A new shuttle route from Bellevue TC to South Bellevue P&R via Bellevue Way.
        4) Move the 560 or 555/6 (if it becomes an all-day route) from 112th to Bellevue Way. Ridership on 112th is about 1/10th of ridership on Bellevue Way, so Bellevue Way wins if you had to pick just one.

        Of those, I think 1 and 4 are the most likely contenders.

    2. I think the strongest argument for creating a route 551 is that the parking at South Bellevue P&R is going away for 8 years while EastLink is under construction. Without the parking, the South Bellevue stop is going to draw much fewer riders than it does today, so the justification of having every single trip serve that stop (and wait 2+ minutes for the left turns) is no longer there.

      Furthermore, a 551 stop at Wilburton Hill P&R would help mitigate the loss of parking at South Bellevue P&R by utilizing an existing parking lot that, today, is pretty under-used due to lack of service.

      1. Wilburton P&R has 99% utilization according to the 2014 Park and Ride Utilization Report. So no, there is not excess capacity there.

        When construction starts, the entire South Bellevue P&R will be closed and so the southbound stop will be moved onto Bellevue Way, eliminating the left turns. I’m not sure what KCM will do about layover space for the 249.

        Current ST service (550, 555/6, and 560) managed to combine for an average of 891 boardings and 862 alightings per day (~750 were on the 550) at South Bellevue in 2013 (from the 2014 SIP). I don’t know what KCM 241 and 249 ridership are, but it likely isn’t zero since the service exists. The lot has 519 stalls (although utilization is 106%, or 551 stalls), so either there is more transfer activity going on than anybody realizes, or there is a not-insignificant number of people accessing the P&R without parking there. Ridership at South Bellevue will likely decline a lot without the parking, but it won’t go to zero and might stay above 200 per day. As an on-the-way stop on the 550, that is still worth serving, especially if alternative service requires riders to backtrack to downtown Bellevue.

  7. why not take 520?

    (runs away)

    I mean, for a BTC -> Cap Hill (Seattle), dgaf-what’s-in-between, wake-me-up-at-Westlake rider, the best choice was to 555 -> 545 / 255. Since this 551 proposal skips both Mercer Island and Rainier Beach Freeway Station, it’s functionally the same as the 520 routing, except with tunnel stops reversed and without the connection.

    Now, I think there could be an argument for a faster BTC -> IDS route, and I don’t think it should skip the Rainier Beach stop.

    Rainier Beach could be a great connection point. If the 9 were actually frequent and fast, 550 -> 9 would be a reasonable Bellevue -> mid-Cap Hill route. (I forget if the FHSC changes anything, but the F doesn’t stand for Fast.) Kiss-and-ride pickups already happen under that stop and is probably the way a lot of southeast Seattle makes use of the 550.

    1. If improving Bellevue-North Capitol Hill connectivity is the desire, you’d be better off working to improve the 271 by deleting the Medina local service. A new all-day Bellevue-Seattle express over SR 520 isn’t in the cards because it would cost too much and duplicate existing service. During peak, the 555/556 facilitate such connections. For those simply going from one endpoint to another (e.g. BTC to Westlake), looking at the travel times for the 556 and 255, using 520 on an express routing would save ~5 minutes over the 550.

      Your point about the Rainier freeway station is well-taken, but service on Rainier is largely limited to the 7. When I lived on Capitol Hill and worked in Issaquah, I did use the 9 –> 554 connection at Rainier sometimes, but the schedules didn’t line up well so the transfer time was marginal. Miss that connection and I would be screwed since the 554 only runs half-hourly at peak, so I tended to go downtown on the 49 instead. Or drive.

      Off-peak (daytime), you could transfer to the 7 and then to the FHSC at Jackson, but the FHSC has 15-minute off-peak frequency and is slow, so you’re probably better off just going downtown and transferring to the 10/11/43/49 instead.

    2. I really think that the frequent, all-day route serving Ranier should be the 9, not the 7, as those who want to go downtown can hop on Link at Mt. Baker Station or take the 36 or First Hill Streetcar from Jackson. The case would have been stronger if they built the 20 feet of elevated sidewalk to connect the pedestrian bridge to the station platform.

      I also realize that the area has a strong vocal minority opposed to any change whatsoever, which would make any proposal to replace the 7 with the 9 DOA.

      Eventually, EastLink and Central Link trains will serve Capitol Hill directly, making this somewhat less of an issue.

  8. >>Jason Rogers says
    February 18, 2015 at 4:34 pm

    The 550 goes away when East Link opens.<<

    Are you just speculating, or do you know this for sure?

    Whatever they do, they need to maintain the same service frequency they have now. You can't go from a bus every 5 minutes during peak times to a bus every 15 minutes and just tell the people along that corridor that their worse service is "for the greater good".

    I think they'll just shorten the route to go from South Bellevue to dt. Bellevue, with the same service frequencies that it has now, requiring about half the buses used now. There's no reason for any more complicated solution.

    1. As far as rerouting the 249 goes, it only runs until like 8pm, and has abysmal service frequency. You can’t just degrade service frequency along that corridor because you built a new rail line elsewhere.

      1. On top of that, the 249 takes a much longer route. People along the Bellevue Way corridor need service to South Bellevue P&R, where they can transfer to East Link, that is just as fast and just as frequent as the service they have now.

    2. CC-

      Sound Transit’s current intention is to delete the 550 when East Link opens. This comes from the mouths of their East Link project team. It doesn’t sound like anyone (at Sound Transit or Metro) has given the issue of what to do about service in that part of Bellevue. As Dan Ryan has noted though, East Link is still 7-8 years away, so a lot can change. As you point out, losing the 550 means the loss of frequent all-day and evening service for that area. How bus service in that area works is one of the things that will need to be addressed; I pointed out what I believe are the two most likely options based on what we know today.

      I mentioned the possibility of a South Bellevue-Bellevue TC shuttle serving Bellevue Way, although this would likely be a Metro route, not a Sound Transit route. Modifications to the 249 would probably include a frequency bump. Under this scenario, it would be faster for riders on Bellevue Way to backtrack to Bellevue TC than to ride the Enatai milk run to South Bellevue.

      Peak service frequency on Bellevue Way is likely to drop, however, as the 5-minute peak frequency is predicated on the massive peak ridership at Bellevue TC, South Bellevue, and Mercer Island, all of which will be handled by East Link. The 550 gets by fine with 10-15 minute frequency off-peak and in the counter-peak direction, so as of today that’s what I would expect for future service along Bellevue Way.

      This is wrapped up in the larger picture of what to do about bus connections to East Link though. Metro and Sound Transit are discussing I-90 route integration via Mercer Island or South Bellevue. There has also been no determination about what to do with the 554 and 555/6; for example, once East Link opens there’s a strong argument for deleting the 554 and making the 555/6 an all-day route with the 554’s span of service (5am-11pm) and increased frequency, and possibly routing it along Bellevue Way as I mentioned above. I consider this the most likely outcome because it solves a variety of transit connectivity issues in a straightforward way.

    3. East Link was always going to replace the 550. Replacement service on Bellevue Way is Metro’s issue, and we’ll probably hear more about it in 2021. Metro’s network then may be different than it is now. I expect more revenue will eventually come to the Eastside, either as a city supplement as Seattle has done, or as a countywide measure, or the legislature becoming more supportive. They really can’t keep their head in the sand and ignore transit forever, especially with the population rising and GenXers and Millenials getting into positions of power. Bellevue’s Transit Master Plan envisions a frequent bus network; that’s only going to happen if the city and county and state get moving.

      So anyway, with the current network the easiest thing to do would be to extend a north Bellevue route to the P&R. It wouldn’t be frequent, but it wasn’t frequent before the 550 either. Bellevue Way isn’t entitled to frequency just because it happens to be an accidental tail of a regional express route. And the 550 is half-hourly evenings and Sundays. But if Bellevue’s Transit Master Plan gets implemented, it would be 15-minute frequent full time all over the place, including probably on Bellevue Way.

  9. Late to the discussion, but I had the EXACT SAME IDEA (well, almost). There are more riders traveling to/from Bellevue than Mercer Island. I propose a new 551 serving the Bellevue T.C., NE 4th & Bellevue Way, S. Bellevue and head nonstop to downtown. Perhaps every other 550 trip during peak hour should be converted? Mercer Island residents can use remaining 550 trips and maybe a few extra, modified 216 trips to commute downtown.

    Bottom line, not every 550 trip needs to deviate to Mercer Island. Every time I’ve used the 550, the bus is mostly full by the time it reaches NE 4th & Bellevue and then becomes standing-room only at S. Bellevue.

Comments are closed.